In [1]:
import pandas as pd
import seaborn as sns
import numpy as np
import matplotlib.pyplot as plt

Exploring the data¶

In [2]:
data=pd.read_csv('S:\Data sets\EVDatabase 2023.csv')
In [3]:
data.head() # checking the top 5 values to get the overview of data.
Out[3]:
Name Subtitle Acceleration TopSpeed Range Efficiency FastChargeSpeed Drive NumberofSeats PriceinGermany PriceinUK
0 Lucid Air Dream Edition P 118 kWh useable battery Available sin... 2.7 sec 270 km/h 645 km 183 Wh/km 820 km/h All Wheel Drive 5 €218,000 NaN
1 Porsche Taycan Turbo S 83.7 kWh useable battery Available si... 2.8 sec 260 km/h 400 km 209 Wh/km 980 km/h All Wheel Drive 4 €189,668 £142,400
2 Audi e-tron GT RS 85 kWh useable battery Available sinc... 3.3 sec 250 km/h 405 km 210 Wh/km 1000 km/h All Wheel Drive 4 €146,050 £115,000
3 Renault Zoe ZE50 R110 52 kWh useable battery Available sinc... 11.4 sec 135 km/h 315 km 165 Wh/km 230 km/h Front Wheel Drive 5 €36,840 NaN
4 Audi Q4 e-tron 35 52 kWh useable battery Available sinc... 9.0 sec 160 km/h 285 km 182 Wh/km 360 km/h Rear Wheel Drive 5 NaN NaN
In [4]:
data.shape
Out[4]:
(309, 11)
In [5]:
data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 309 entries, 0 to 308
Data columns (total 11 columns):
 #   Column           Non-Null Count  Dtype 
---  ------           --------------  ----- 
 0   Name             309 non-null    object
 1   Subtitle         309 non-null    object
 2   Acceleration     309 non-null    object
 3   TopSpeed         309 non-null    object
 4   Range            309 non-null    object
 5   Efficiency       309 non-null    object
 6   FastChargeSpeed  309 non-null    object
 7   Drive            309 non-null    object
 8   NumberofSeats    309 non-null    int64 
 9   PriceinGermany   282 non-null    object
 10  PriceinUK        198 non-null    object
dtypes: int64(1), object(10)
memory usage: 26.7+ KB
In [6]:
data.isna().sum() #counting the null values in each column
Out[6]:
Name                 0
Subtitle             0
Acceleration         0
TopSpeed             0
Range                0
Efficiency           0
FastChargeSpeed      0
Drive                0
NumberofSeats        0
PriceinGermany      27
PriceinUK          111
dtype: int64
In [7]:
sns.heatmap(data.isna()) # visually checking the null vlues
Out[7]:
<Axes: >
In [8]:
# Observing the Name column we conclude that first word in the string is the name of the company followed by model name
# The brand valuse of the company also affects the marketprice

Cleaning the data¶

Creating a column for Brand¶

In [9]:
data.columns
Out[9]:
Index(['Name', 'Subtitle', 'Acceleration', 'TopSpeed', 'Range', 'Efficiency',
       'FastChargeSpeed', 'Drive', 'NumberofSeats', 'PriceinGermany',
       'PriceinUK'],
      dtype='object')
In [10]:
#creating a new dataframe 'df' to store Brand name 
df = pd.DataFrame({'Brand': data['Name'].str.split().str[0]}) #making dictionary to store values
df
Out[10]:
Brand
0 Lucid
1 Porsche
2 Audi
3 Renault
4 Audi
... ...
304 Volkswagen
305 Volkswagen
306 Polestar
307 Polestar
308 Maserati

309 rows × 1 columns

Understanding the content in 'Subtitle'¶

In [11]:
data.Subtitle.value_counts()
Out[11]:
68 kWh useable battery          Available since February 2021       8
46.3 kWh useable battery          Available since November 2020     8
77 kWh useable battery          Available since March 2022          5
46.3 kWh useable battery          Available since November 2021     5
50.8 kWh useable battery          Expected from May 2023            4
                                                                   ..
49 kWh useable battery          Available since November 2021       1
107.8 kWh useable battery          Available since February 2022    1
40 kWh useable battery          Available since April 2022          1
74 kWh useable battery          Available since September 2021      1
95 kWh useable battery          Expected from September 2023        1
Name: Subtitle, Length: 187, dtype: int64

Extracting the battery power data¶

In [12]:
# There are two set of sentences 
# 1st sentence is about battery 
# 2nd sentence is about availability of the battery in market
In [13]:
df['Batery_KWh']=data.Subtitle.str.split().str[0]
df['Batery_KWh'].head() # 'Batery_KWh' will have the value of battery power allocated to vehicle
Out[13]:
0     118
1    83.7
2      85
3      52
4      52
Name: Batery_KWh, dtype: object
In [14]:
data.Subtitle.head() #Comparying the top 5 values for verification
Out[14]:
0    118 kWh useable battery          Available sin...
1    83.7 kWh useable battery          Available si...
2    85 kWh useable battery          Available sinc...
3    52 kWh useable battery          Available sinc...
4    52 kWh useable battery          Available sinc...
Name: Subtitle, dtype: object
Understanding the term KWh¶

KWh stands for kilowatt hour, and it is a unit of energy. It is the amount of energy used by a device that consumes one kilowatt of power for one hour. In the context of electric vehicles (EVs), kWh is used to measure the capacity of the battery and the energy consumption of the vehicle. The battery capacity of an EV is the amount of energy that the battery can store. It is typically measured in kWh. The higher the battery capacity, the longer the EV can travel on a single charge. The energy consumption of an EV is the amount of energy that the vehicle uses to travel a certain distance. It is typically measured in kWh/100km. The lower the energy consumption, the more efficient the EV is.

In [15]:
data.head(2)
Out[15]:
Name Subtitle Acceleration TopSpeed Range Efficiency FastChargeSpeed Drive NumberofSeats PriceinGermany PriceinUK
0 Lucid Air Dream Edition P 118 kWh useable battery Available sin... 2.7 sec 270 km/h 645 km 183 Wh/km 820 km/h All Wheel Drive 5 €218,000 NaN
1 Porsche Taycan Turbo S 83.7 kWh useable battery Available si... 2.8 sec 260 km/h 400 km 209 Wh/km 980 km/h All Wheel Drive 4 €189,668 £142,400

Removing the units from other columns¶

In [16]:
df['accln_sec']= data['Acceleration'].str.split().str[0]
In [17]:
df['TopSpeed_km/h']= data['TopSpeed'].str.split().str[0]
In [18]:
df['range_km']= data['Range'].str.split().str[0]
In [19]:
df['efficiency_wh/km']= data['Efficiency'].str.split().str[0]
In [20]:
df['ChargeSpeed_km/hr']= data['FastChargeSpeed'].str.split().str[0]
In [21]:
df['drive']= data['Drive']
In [22]:
df['NumberofSeats']= data['NumberofSeats']
In [23]:
df['PriceinGermany']= data['PriceinGermany'].str.replace('€','',regex=True)
In [24]:
df['PriceinGermany']= df['PriceinGermany'].str.replace(',','',regex=True)
In [25]:
df['PriceinUK']= data['PriceinUK'].str.replace('£','',regex=True)
In [26]:
df['PriceinUK']= df['PriceinUK'].str.replace(',','',regex=True)

Checking the new data frame 'df'¶

In [27]:
df.head()
Out[27]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
0 Lucid 118 2.7 270 645 183 820 All Wheel Drive 5 218000 NaN
1 Porsche 83.7 2.8 260 400 209 980 All Wheel Drive 4 189668 142400
2 Audi 85 3.3 250 405 210 1000 All Wheel Drive 4 146050 115000
3 Renault 52 11.4 135 315 165 230 Front Wheel Drive 5 36840 NaN
4 Audi 52 9.0 160 285 182 360 Rear Wheel Drive 5 NaN NaN
In [28]:
df.shape
Out[28]:
(309, 11)
In [29]:
df.info() # checking the dtypes of data frame df

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 309 entries, 0 to 308
Data columns (total 11 columns):
 #   Column             Non-Null Count  Dtype 
---  ------             --------------  ----- 
 0   Brand              309 non-null    object
 1   Batery_KWh         309 non-null    object
 2   accln_sec          309 non-null    object
 3   TopSpeed_km/h      309 non-null    object
 4   range_km           309 non-null    object
 5   efficiency_wh/km   309 non-null    object
 6   ChargeSpeed_km/hr  309 non-null    object
 7   drive              309 non-null    object
 8   NumberofSeats      309 non-null    int64 
 9   PriceinGermany     282 non-null    object
 10  PriceinUK          198 non-null    object
dtypes: int64(1), object(10)
memory usage: 26.7+ KB
In [30]:
#The columns 'PriceinGermany' and 'PriceinUK' should be intiger.
In [31]:
# the blank spaces or null values have to be filled first before changing data types.
# filling the '0' in place null values 
df['PriceinUK'] = df['PriceinUK'].fillna(0) 
df['PriceinGermany']= df['PriceinGermany'].fillna(0)
# Changing the datatype from object to int
df['PriceinUK']= df['PriceinUK'].astype(int)
df['PriceinGermany']= df['PriceinGermany'].astype(int)
In [32]:
#changine the data types of remaining columns
df['Batery_KWh'] = df['Batery_KWh'].astype(float) #this columns have values in decimal so we will uconvert to float
df['accln_sec'] = df['accln_sec'].astype(float) #this columns have values in decimal so we will uconvert to float
df['range_km'] = df['range_km'].astype(int)
df['efficiency_wh/km'] = df['efficiency_wh/km'].astype(int)
df['TopSpeed_km/h'] = df['TopSpeed_km/h'].astype(int)
In [33]:
df['ChargeSpeed_km/hr'].unique() # there is '-' value which needs to be fixed for changing the data type
Out[33]:
array(['820', '980', '1000', '230', '360', '640', '1050', '1020', '-',
       '180', '970', '1130', '310', '150', '730', '380', '440', '530',
       '500', '540', '390', '370', '430', '520', '840', '960', '290',
       '280', '1010', '1170', '790', '770', '650', '630', '260', '160',
       '1030', '220', '950', '480', '680', '330', '720', '750', '1070',
       '900', '510', '490', '470', '460', '570', '670', '320', '400',
       '780', '410', '190', '560', '1110', '420', '300', '710', '930',
       '1090', '910', '1060', '450', '990', '170', '240', '760', '210',
       '870', '920', '880', '340', '1240', '1150', '860', '580', '350',
       '200', '610', '700', '940', '600', '590', '550', '620', '850',
       '800'], dtype=object)
In [34]:
df['ChargeSpeed_km/hr'] = df['ChargeSpeed_km/hr'].str.replace('-', '').str[0].fillna(0).astype(int)
# temporarily changing '-' with 0
In [35]:
df.dtypes # checking the datatypes
Out[35]:
Brand                 object
Batery_KWh           float64
accln_sec            float64
TopSpeed_km/h          int32
range_km               int32
efficiency_wh/km       int32
ChargeSpeed_km/hr      int32
drive                 object
NumberofSeats          int64
PriceinGermany         int32
PriceinUK              int32
dtype: object
In [36]:
df.head()
Out[36]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
0 Lucid 118.0 2.7 270 645 183 8 All Wheel Drive 5 218000 0
1 Porsche 83.7 2.8 260 400 209 9 All Wheel Drive 4 189668 142400
2 Audi 85.0 3.3 250 405 210 1 All Wheel Drive 4 146050 115000
3 Renault 52.0 11.4 135 315 165 2 Front Wheel Drive 5 36840 0
4 Audi 52.0 9.0 160 285 182 3 Rear Wheel Drive 5 0 0
In [37]:
df[df['ChargeSpeed_km/hr']==0]
Out[37]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
9 Smart 16.7 11.6 130 100 167 0 Rear Wheel Drive 2 21940 0
10 Smart 16.7 11.9 130 95 176 0 Rear Wheel Drive 2 25200 0
17 Renault 21.3 12.6 135 130 164 0 Rear Wheel Drive 4 28000 0
In [38]:
df['ChargeSpeed_km/hr']=df.groupby('Brand')['ChargeSpeed_km/hr'].transform(lambda x:x.replace(0,x.mean()))
In [39]:
df[df['ChargeSpeed_km/hr']==0] # Null values fixed with mean
Out[39]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK

Checking and counting the unique fields in dataframe¶

In [40]:
df.Brand.unique()
Out[40]:
array(['Lucid', 'Porsche', 'Audi', 'Renault', 'Tesla', 'Smart', 'Honda',
       'Mercedes', 'Lexus', 'BMW', 'Fiat', 'Skoda', 'Nissan',
       'Volkswagen', 'Citroen', 'Opel', 'Peugeot', 'Mini', 'JAC',
       'Hyundai', 'Kia', 'MG', 'Toyota', 'CUPRA', 'Subaru', 'SsangYong',
       'Genesis', 'Aiways', 'Mazda', 'Dacia', 'Fisker', 'Hongqi', 'NIO',
       'Ford', 'Polestar', 'Rolls-Royce', 'Lotus', 'Volvo', 'ORA', 'BYD',
       'Abarth', 'DS', 'Jeep', 'Maserati', 'Seres', 'VinFast', 'Jaguar',
       'XPENG', 'Maxus'], dtype=object)
In [41]:
df.Brand.nunique() #counting total brands
Out[41]:
49
In [42]:
df.Brand.value_counts()
Out[42]:
Mercedes       38
Porsche        18
Audi           16
Peugeot        15
Volkswagen     15
Opel           14
Citroen        12
BMW            10
Tesla          10
Toyota         10
Hyundai        10
Fiat           10
Kia             9
Skoda           8
Polestar        8
Volvo           8
MG              8
Renault         7
Nissan          6
NIO             6
XPENG           6
VinFast         6
Genesis         5
Ford            5
Lucid           5
Fisker          4
Smart           4
CUPRA           3
ORA             3
BYD             3
Aiways          2
Dacia           2
Hongqi          2
Mini            2
Lexus           2
Maserati        2
Lotus           2
Jeep            2
DS              1
Seres           1
Jaguar          1
Subaru          1
Abarth          1
Rolls-Royce     1
Mazda           1
SsangYong       1
JAC             1
Honda           1
Maxus           1
Name: Brand, dtype: int64
In [43]:
df.drive.nunique()
Out[43]:
3
In [44]:
df.drive.value_counts()
Out[44]:
All Wheel Drive      128
Front Wheel Drive    115
Rear Wheel Drive      66
Name: drive, dtype: int64
In [45]:
df.NumberofSeats.nunique()
Out[45]:
4
In [46]:
df.NumberofSeats.value_counts()
Out[46]:
5    203
7     66
4     38
2      2
Name: NumberofSeats, dtype: int64
In [47]:
# Checking if there is a difference between 'PriceinGermany' and 'PriceinUK' because of difference in currency value wrt dollar
In [48]:
(df.PriceinGermany - df.PriceinUK).value_counts()
Out[48]:
 0         18
 5000       6
 12990      4
 7000       3
 7495       3
           ..
 209        1
 28214      1
 28166      1
 1780       1
-135000     1
Length: 263, dtype: int64

The price differences are not due to difference between the strengths of curriencies.

In [49]:
df.describe().T # analysing the values
Out[49]:
count mean std min 25% 50% 75% max
Batery_KWh 309.0 71.076052 21.139409 16.7 56.0 74.0 85.0 123.0
accln_sec 309.0 7.532039 3.127109 2.1 4.9 7.0 9.0 19.1
TopSpeed_km/h 309.0 180.754045 38.738187 125.0 150.0 180.0 200.0 320.0
range_km 309.0 361.440129 110.861212 95.0 275.0 370.0 440.0 685.0
efficiency_wh/km 309.0 199.430421 33.021626 150.0 174.0 192.0 214.0 295.0
ChargeSpeed_km/hr 309.0 4.295192 2.249713 1.0 2.0 4.0 6.0 9.0
NumberofSeats 309.0 5.284790 0.978567 2.0 5.0 5.0 5.0 7.0
PriceinGermany 309.0 64832.462783 44638.555529 0.0 42000.0 56942.0 72519.0 400000.0
PriceinUK 309.0 41173.233010 43035.103703 0.0 0.0 37000.0 57990.0 350000.0

The min value is represented by 0 bcz we changed the NaN values to 0 in above operations. We will have to replace 0 with mean values.

Fixing the null value¶

In [50]:
# 
df['PriceinGermany']=df.groupby('Brand')['PriceinGermany'].transform(lambda x:x.replace(0,x.mean()))
In [51]:
(df['PriceinGermany']==0).sum() # earlier there were 27 Null values
Out[51]:
11
In [52]:
df['PriceinUK']=df.groupby('Brand')['PriceinUK'].transform(lambda x:x.replace(0,x.mean()))
In [53]:
(df['PriceinUK']==0).sum() # earlier there were 111 Null values
Out[53]:
32
In [54]:
df[df['PriceinGermany']==0].head(11) # this data can be used for prediction
Out[54]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
56 JAC 39.0 12.0 132 225 173 1.0 Front Wheel Drive 5 0.0 0.0
214 Hongqi 76.5 6.5 200 305 251 4.0 All Wheel Drive 5 0.0 0.0
215 Hongqi 90.0 4.9 200 355 254 4.0 All Wheel Drive 5 0.0 0.0
271 Seres 51.0 8.9 155 260 196 2.0 Front Wheel Drive 5 0.0 0.0
287 XPENG 82.7 6.7 200 500 165 7.0 Rear Wheel Drive 5 0.0 0.0
288 XPENG 82.7 4.1 200 470 176 6.0 All Wheel Drive 5 0.0 0.0
289 XPENG 82.7 4.1 200 465 178 6.0 All Wheel Drive 5 0.0 0.0
290 XPENG 75.0 6.4 200 375 200 7.0 Rear Wheel Drive 5 0.0 0.0
291 XPENG 94.0 6.4 200 470 200 9.0 Rear Wheel Drive 5 0.0 0.0
292 XPENG 94.0 3.9 200 440 214 9.0 All Wheel Drive 5 0.0 0.0
302 Maxus 90.0 9.2 180 345 261 2.0 Front Wheel Drive 7 0.0 64306.0
In [55]:
case1=df[df['PriceinGermany']==0]
In [56]:
case1.shape
Out[56]:
(11, 11)
In [57]:
case1.head()
Out[57]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
56 JAC 39.0 12.0 132 225 173 1.0 Front Wheel Drive 5 0.0 0.0
214 Hongqi 76.5 6.5 200 305 251 4.0 All Wheel Drive 5 0.0 0.0
215 Hongqi 90.0 4.9 200 355 254 4.0 All Wheel Drive 5 0.0 0.0
271 Seres 51.0 8.9 155 260 196 2.0 Front Wheel Drive 5 0.0 0.0
287 XPENG 82.7 6.7 200 500 165 7.0 Rear Wheel Drive 5 0.0 0.0
In [58]:
df = df[df['PriceinGermany'] != 0] # Removal of null values
In [59]:
case2=df[df['PriceinUK']==0]
In [60]:
case2.shape
Out[60]:
(22, 11)
In [61]:
case2.head() #will be used for prediction
Out[61]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
0 Lucid 118.0 2.7 270 645 183 8.0 All Wheel Drive 5 218000.0 0.0
31 Lucid 112.0 3.2 270 665 168 8.0 All Wheel Drive 5 174500.0 0.0
32 Lucid 88.0 3.4 250 550 160 9.0 All Wheel Drive 5 120000.0 0.0
33 Lucid 88.0 4.2 200 560 157 9.0 Rear Wheel Drive 5 100000.0 0.0
122 SsangYong 56.0 8.5 156 290 193 3.0 Front Wheel Drive 5 40490.0 0.0
In [62]:
df = df[df['PriceinUK'] != 0]# Removal of null values
In [63]:
df.shape
Out[63]:
(276, 11)

Insights of the extremes of each column¶

In [64]:
# Most and least luxurious cars based on 'PriceinGermany'
most_luxurious = df.sort_values(by='PriceinGermany', ascending=False).head(1)
most_affordable = df.sort_values(by='PriceinGermany').head(1)
In [65]:
most_luxurious
Out[65]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
233 Rolls-Royce 100.0 4.5 250 455 220 5.0 All Wheel Drive 4 400000.0 350000.0
In [66]:
most_affordable
Out[66]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
9 Smart 16.7 11.6 130 100 167 2.25 Rear Wheel Drive 2 21940.0 19250.0
In [67]:
# Most and least powerful battery cars based on 'Batery_KWh'
most_powerful_battery = df.sort_values(by='Batery_KWh', ascending=False).head(1)
least_powerful_battery = df.sort_values(by='Batery_KWh').head(1)
In [68]:
most_powerful_battery
Out[68]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
175 Mercedes 108.4 4.6 210 485 224 6.0 All Wheel Drive 7 135434.0 139170.0
In [69]:
least_powerful_battery
Out[69]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
9 Smart 16.7 11.6 130 100 167 2.25 Rear Wheel Drive 2 21940.0 19250.0
In [70]:
# Fastest and slowest acceleration cars based on 'accln_sec'
fastest_acceleration = df.sort_values(by='accln_sec').head(1)
slowest_acceleration = df.sort_values(by='accln_sec', ascending=False).head(1)
In [71]:
fastest_acceleration
Out[71]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
52 Tesla 95.0 2.1 282 550 173 7.0 All Wheel Drive 5 137990.0 125000.0
In [72]:
slowest_acceleration
Out[72]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
48 Peugeot 68.0 14.3 130 260 262 2.0 Front Wheel Drive 7 60430.0 23311.666667
In [73]:
# Fastest and slowest cars based on 'TopSpeed_km/h'
fastest_car = df.sort_values(by='TopSpeed_km/h').head(1)
slowest_car = df.sort_values(by='TopSpeed_km/h', ascending=False).head(1)
In [74]:
fastest_car
Out[74]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
127 Citroen 68.0 13.3 130 265 257 2.0 Front Wheel Drive 7 57940.0 21883.75
In [75]:
slowest_car
Out[75]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
270 Maserati 83.0 2.7 320 425 195 9.0 All Wheel Drive 4 250000.0 200000.0
In [76]:
# Cars with the longest and shortest range based on 'range_km'
longest_range = df.sort_values(by='range_km', ascending=False).head(1)
shortest_range = df.sort_values(by='range_km').head(1)
In [77]:
longest_range
Out[77]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
73 Mercedes 107.8 6.2 210 640 168 9.0 Rear Wheel Drive 5 109551.0 105610.0
In [78]:
shortest_range
Out[78]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
10 Smart 16.7 11.9 130 95 176 2.25 Rear Wheel Drive 2 25200.0 19250.0
In [79]:
# Cars with the highest and lowest efficiency based on 'efficiency_wh/km'
highest_efficiency = df.sort_values(by='efficiency_wh/km').head(1)
lowest_efficiency = df.sort_values(by='efficiency_wh/km', ascending=False).head(1)
In [80]:
highest_efficiency
Out[80]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
200 Hyundai 54.0 8.8 185 360 150 8.0 Rear Wheel Drive 5 43900.0 38214.0
In [81]:
lowest_efficiency
Out[81]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
30 Mercedes 90.0 12.1 160 305 295 3.0 Front Wheel Drive 7 72519.0 44560.078947
In [82]:
# Cars with the fastest and slowest charge speed based on 'ChargeSpeed_km/hr'
fastest_charge_speed = df.sort_values(by='ChargeSpeed_km/hr').head(1)
slowest_charge_speed = df.sort_values(by='ChargeSpeed_km/hr', ascending=False).head(1)
In [83]:
fastest_charge_speed
Out[83]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
49 Porsche 71.0 5.4 230 410 173 1.0 Rear Wheel Drive 4 88399.0 75500.0
In [84]:
slowest_charge_speed
Out[84]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
1 Porsche 83.7 2.8 260 400 209 9.0 All Wheel Drive 4 189668.0 142400.0
In [85]:
# Cars with the most and least number of seats based on 'NumberofSeats'
most_seats = df.sort_values(by='NumberofSeats', ascending=False).head(10)
least_seats = df.sort_values(by='NumberofSeats').head(2)
In [86]:
most_seats
Out[86]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
144 Mercedes 60.0 12.0 160 205 293 3.0 Front Wheel Drive 7 61571.0 44560.078947
126 Citroen 46.3 12.1 130 180 257 2.0 Front Wheel Drive 7 52730.0 35995.000000
110 Mercedes 66.5 8.0 160 335 199 4.0 All Wheel Drive 7 55519.0 55310.000000
105 Citroen 46.3 11.7 135 200 232 3.0 Front Wheel Drive 7 43640.0 32495.000000
204 Fiat 46.3 12.1 130 185 250 2.0 Front Wheel Drive 7 55990.0 6429.000000
103 Opel 46.3 11.7 135 200 232 3.0 Front Wheel Drive 7 44750.0 34635.000000
102 Opel 46.3 11.7 135 205 226 3.0 Front Wheel Drive 7 43050.0 34035.000000
101 Mercedes 60.0 12.0 160 205 293 1.0 Front Wheel Drive 7 68949.0 44560.078947
100 Mercedes 60.0 12.0 160 205 293 1.0 Front Wheel Drive 7 68056.0 44560.078947
205 Fiat 68.0 13.3 130 265 257 2.0 Front Wheel Drive 7 61990.0 6429.000000
In [87]:
least_seats
Out[87]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
9 Smart 16.7 11.6 130 100 167 2.25 Rear Wheel Drive 2 21940.0 19250.0
10 Smart 16.7 11.9 130 95 176 2.25 Rear Wheel Drive 2 25200.0 19250.0
In [88]:
# Most and least expensive cars in the UK based on 'PriceinUK'
most_expensive_uk = df.sort_values(by='PriceinUK', ascending=False).head(1)
least_expensive_uk = df.sort_values(by='PriceinUK').head(1)
In [89]:
most_expensive_uk
Out[89]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
233 Rolls-Royce 100.0 4.5 250 455 220 5.0 All Wheel Drive 4 400000.0 350000.0
In [90]:
least_expensive_uk
Out[90]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
207 Fiat 68.0 13.3 130 260 262 2.0 Front Wheel Drive 7 62990.0 6429.0
In [ ]:
In [ ]:

Exploratory Data Analysis¶

Vidualising the Data¶

In [91]:
plt.figure(figsize=(8,6))
sns.heatmap(df.corr(), annot=True)

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\249308572.py:2: FutureWarning: The default value of numeric_only in DataFrame.corr is deprecated. In a future version, it will default to False. Select only valid columns or specify the value of numeric_only to silence this warning.
  sns.heatmap(df.corr(), annot=True)
Out[91]:
<Axes: >

1) Battery capacity is positively correlated with range (0.89) and efficiency (0.62). This means that cars with larger batteries can travel further on a single charge and are more efficient at using energy.

2) Acceleration is negatively correlated with range (-0.75) and efficiency (-0.87). This means that cars that can accelerate faster tend to have shorter ranges and are less efficient.

3) Top speed is also negatively correlated with range (-0.50) and efficiency (-0.75). This means that cars with higher top speeds tend to have shorter ranges and are less efficient.

4) Price is positively correlated with battery capacity (0.62), range (0.51), efficiency (0.75), and top speed (0.72). This means that more expensive cars typically have larger batteries, longer ranges, are more efficient, and have higher top speeds.

In [92]:
df.columns
Out[92]:
Index(['Brand', 'Batery_KWh', 'accln_sec', 'TopSpeed_km/h', 'range_km',
       'efficiency_wh/km', 'ChargeSpeed_km/hr', 'drive', 'NumberofSeats',
       'PriceinGermany', 'PriceinUK'],
      dtype='object')
In [93]:
ax=df.Brand.value_counts()  #visualising the available options to choose from
In [94]:
plt.figure(figsize=(12,5))

bx=sns.barplot(x=ax.index, y=ax.values)


plt.title('Number of electric cars offered by each company')
plt.xlabel('Car Brand')
plt.ylabel('Count')
plt.xticks(rotation=90)

for bars in bx.containers:
    bx.bar_label(bars)

Mercedes offer most number of electric vehicles in the market.

In [95]:
ax=df.drive.value_counts() #Distribution of drive systems
In [96]:
plt.figure(figsize=(4, 4))
plt.pie(ax, labels=ax.index, autopct='%1.1f%%', startangle=90)
plt.title('Popular Drive systems')
Out[96]:
Text(0.5, 1.0, 'Popular Drive systems')

Majority vehicles are using all wheel drive and front wheel drive.

In [97]:
plt.figure(figsize=(7,5)) #visualizing the count of the drive system 

ax=sns.countplot(x='drive', data=df)

plt.title('Drive system Distribution and count')
plt.xlabel('drive system')
plt.ylabel('Count')
plt.xticks(rotation=0)

for bars in ax.containers:
    ax.bar_label(bars)
In [98]:
plt.figure(figsize=(10,18)) # 

ax=sns.countplot(y='Brand', hue='drive', data=df)

plt.title('Drive system used by manufacturers')
plt.ylabel('Car Brand')
plt.xlabel('Count')
plt.xticks(rotation=0)

for bars in ax.containers:
    ax.bar_label(bars)
In [99]:
plt.figure(figsize=(7,5)) # 

ax=sns.countplot(x='NumberofSeats', data=df)

plt.title('Seat distributionan')
plt.xlabel('seats')
plt.ylabel('Count')
plt.xticks(rotation=0)

for bars in ax.containers:
    ax.bar_label(bars)

Majority of vehicles are having 5 seats.

In [100]:
plt.figure(figsize=(7,5)) # 

ax=sns.countplot(x='NumberofSeats', hue='drive', data=df)

plt.title('Seat distributionan and drive')
plt.xlabel('seats')
plt.ylabel('Count')
plt.xticks(rotation=0)

for bars in ax.containers:
    ax.bar_label(bars)

1) Majority od vehicles having 5 seats use all wheel drive. 2) Front wheel drive is preferred more in 7 seater vehicles.

In [101]:
# checking if brandvalue have an effect on cost or not.
brand_cost=df.groupby(['Brand'],as_index=False)['PriceinGermany'].mean().sort_values(by='PriceinGermany',ascending=False)
In [102]:
brand_cost.head()
Out[102]:
Brand PriceinGermany
30 Rolls-Royce 400000.000000
19 Maserati 187500.000000
28 Porsche 130015.944444
17 Lotus 123490.000000
2 BMW 95820.000000
In [103]:
sns.relplot(y="Brand", x="PriceinGermany", height=10, aspect=0.8 ,data=brand_cost)

plt.title('Average cost of an EV for each brand')
plt.xlabel('Average price in Germany')
plt.ylabel('Brand')
Out[103]:
Text(-50.55555555555555, 0.5, 'Brand')
In [104]:
# checking if brandvalue have an effect on cost or not.
brand_cost_uk =df.groupby(['Brand'],as_index=False)['PriceinUK'].mean().sort_values(by='PriceinUK',ascending=False)
In [105]:
brand_cost_uk.head()
Out[105]:
Brand PriceinUK
30 Rolls-Royce 350000.000000
19 Maserati 167500.000000
17 Lotus 104750.000000
28 Porsche 102628.333333
2 BMW 86985.000000
In [106]:
sns.relplot(y="Brand", x="PriceinUK", height=10, aspect=0.8 ,data=brand_cost_uk)

plt.title('Average cost of an EV for each brand')
plt.xlabel('Average price in UK')
plt.ylabel('Brand')
Out[106]:
Text(-50.55555555555555, 0.5, 'Brand')
In [107]:
#based on usage there will be relation between the battery capacity and  acceleration of the vehicle.
sns.relplot(x="Batery_KWh", y="accln_sec", height=6,aspect=2,hue="drive",data=df)

plt.title('Battery capacity vs acceleration')
plt.xlabel('Battery  capacity (kwh)')
plt.ylabel('acceleration (sec)')
plt.xticks(rotation=0)
Out[107]:
(array([  0.,  20.,  40.,  60.,  80., 100., 120.]),
 [Text(0.0, 0, '0'),
  Text(20.0, 0, '20'),
  Text(40.0, 0, '40'),
  Text(60.0, 0, '60'),
  Text(80.0, 0, '80'),
  Text(100.0, 0, '100'),
  Text(120.0, 0, '120')])

All wheel drive cars are having huge batteries and the acceleraton is really fast.

In [108]:
sns.relplot(x="Batery_KWh", y="accln_sec", height=6,aspect=2,hue="NumberofSeats",data=df,palette=["red", "green", "blue", "black"])

plt.title('Battery capacity vs acceleration')
plt.xlabel('Battery  capacity (kwh)')
plt.ylabel('acceleration (sec)')
plt.xticks(rotation=0)
Out[108]:
(array([  0.,  20.,  40.,  60.,  80., 100., 120.]),
 [Text(0.0, 0, '0'),
  Text(20.0, 0, '20'),
  Text(40.0, 0, '40'),
  Text(60.0, 0, '60'),
  Text(80.0, 0, '80'),
  Text(100.0, 0, '100'),
  Text(120.0, 0, '120')])

Majority cars are 5 seaters. 5 seater cars which are fast have an all wheel drive system.

In [109]:
#relation of battery and top speed w.r.t. Number of seets

sns.relplot(x="Batery_KWh", y="TopSpeed_km/h", height=6,aspect=2,hue="NumberofSeats",data=df,palette=["red", "green", "blue", "black"])

plt.title('Battery capacity vs top speed')
plt.xlabel('Battery  capacity (kwh)')
plt.ylabel('Top speed (km/h)')
plt.xticks(rotation=0)
Out[109]:
(array([  0.,  20.,  40.,  60.,  80., 100., 120.]),
 [Text(0.0, 0, '0'),
  Text(20.0, 0, '20'),
  Text(40.0, 0, '40'),
  Text(60.0, 0, '60'),
  Text(80.0, 0, '80'),
  Text(100.0, 0, '100'),
  Text(120.0, 0, '120')])
In [110]:
#relation of battery and top speed w.r.t. drive

sns.relplot(x="Batery_KWh", y="TopSpeed_km/h", height=6,aspect=2,hue="drive",data=df,)

plt.title('Battery capacity vs top speed')
plt.xlabel('Battery  capacity (kwh)')
plt.ylabel('Top speed (km/h)')
plt.xticks(rotation=0)
Out[110]:
(array([  0.,  20.,  40.,  60.,  80., 100., 120.]),
 [Text(0.0, 0, '0'),
  Text(20.0, 0, '20'),
  Text(40.0, 0, '40'),
  Text(60.0, 0, '60'),
  Text(80.0, 0, '80'),
  Text(100.0, 0, '100'),
  Text(120.0, 0, '120')])

All wheel drive cars are majoritily 5 seater cars and have better performance in terms of battery capacity and top speed.

In [111]:
# Speed vs range graph
sns.relplot(x="TopSpeed_km/h", y="range_km",height=6, hue="drive",data=df)

plt.title('top speed vs range')
plt.xlabel('top speed (km/h)')
plt.ylabel('range (km)')
Out[111]:
Text(49.62089120370371, 0.5, 'range (km)')
In [112]:
# Speed vs range graph
sns.relplot(x="TopSpeed_km/h", y="range_km",height=6, hue="NumberofSeats",palette=["red", "green", "blue", "black"],data=df)

plt.title('top speed vs range')
plt.xlabel('top speed (km/h)')
plt.ylabel('range (km)')
Out[112]:
Text(43.016428755144034, 0.5, 'range (km)')

All wheel drive EVs are better in terms of renge and top speed.

In [113]:
# Usually fast cars have higher price range so we will plot a graph for top speed and price 

sns.relplot(x="TopSpeed_km/h", y="PriceinGermany",height=6, hue="drive",data=df)

plt.title('top speed vs price')
plt.xlabel('top speed (km/h)')
plt.ylabel('Price in Germany')
Out[113]:
Text(57.364328703703706, 0.5, 'Price in Germany')
In [114]:
# Usually fast cars have higher price range so we will plot a graph for top speed and price 

sns.relplot(x="TopSpeed_km/h", y="PriceinGermany",height=6, hue="NumberofSeats",data=df, palette=["red", "green", "blue", "black"])

plt.title('top speed vs price')
plt.xlabel('top speed (km/h)')
plt.ylabel('Price in Germany')
Out[114]:
Text(48.27979681069958, 0.5, 'Price in Germany')

The faster the car, the higher is the price.

In [115]:
import plotly.express as px # For 3d visualization
In [116]:
fig = px.scatter_3d(df, x='accln_sec', y='efficiency_wh/km', z='PriceinGermany', color='Brand', height=800, width=800)
fig.update_layout(title='3D scatter plot of electric vehicles')
fig.show()

Price, acceleration and efficiency are the major factors considered while a customer tries to buy a car. Mercedes presents a wide range for selection where we can observe that the company has most economical car along with cars that are fast and fun to drive but little less on efficiency.

Conclusion of EDA¶

  1. EV manufacturers either use all wheel drive system or front wheel drive system. Front wheel drive system is available in very limited models.
  2. Cars using all wheel drive systems are equipped with heavy batteries and also seemed to have better efficiency.
  3. Seating capacities ranges from 2 seater upto 7 seater in the data available. The 2 seater cars can be called as concept cars or maybe very old models since they tend to fall on a low end in terms of performance parameters.
  4. All wheel drive cars majoritily have seating capacity of 5 passengers followed by 7 paggengers.
  5. 7 seater cars have generally all wheel drive followed by a front wheel drive.
  6. All wheel drive EVs are comparitively faster and also they are higher on prise range.
  7. 4 seater cars have an optimised speed to range ratio.
  8. 7 seater EVs have low speed and good range.

Predictive modeling¶

In [117]:
df.head()
Out[117]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
1 Porsche 83.7 2.8 260 400 209 9.0 All Wheel Drive 4 189668.000 142400.000
2 Audi 85.0 3.3 250 405 210 1.0 All Wheel Drive 4 146050.000 115000.000
3 Renault 52.0 11.4 135 315 165 2.0 Front Wheel Drive 5 36840.000 9570.000
4 Audi 52.0 9.0 160 285 182 3.0 Rear Wheel Drive 5 70905.625 56860.625
5 Tesla 75.0 3.7 250 415 181 6.0 All Wheel Drive 5 63667.000 59990.000
In [118]:
df.drive.unique()
Out[118]:
array(['All Wheel Drive', 'Front Wheel Drive', 'Rear Wheel Drive'],
      dtype=object)

For the predictive model we will convert the drive column to numeric values with the help of mapping function. The mapping values are as following - 1) Front Wheel Drive: 1 2) Rear Wheel Drive: 2 3) All Wheel Drive: 3

In [119]:
# Define the mapping dictionary
drive_mapping = {'Front Wheel Drive': 1,'Rear Wheel Drive': 2, 'All Wheel Drive': 3 }

# Use the map function to create a new column with the mapped values
df['drive'] = df['drive'].map(drive_mapping)
In [120]:
df.head() #checking the
Out[120]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
1 Porsche 83.7 2.8 260 400 209 9.0 3 4 189668.000 142400.000
2 Audi 85.0 3.3 250 405 210 1.0 3 4 146050.000 115000.000
3 Renault 52.0 11.4 135 315 165 2.0 1 5 36840.000 9570.000
4 Audi 52.0 9.0 160 285 182 3.0 2 5 70905.625 56860.625
5 Tesla 75.0 3.7 250 415 181 6.0 3 5 63667.000 59990.000

Observing the distribution of data¶

In [121]:
df.describe().T
Out[121]:
count mean std min 25% 50% 75% max
Batery_KWh 276.0 69.996377 20.652600 16.7 52.00 70.5 83.700 108.4
accln_sec 276.0 7.691667 3.058515 2.1 5.10 7.3 9.425 14.3
TopSpeed_km/h 276.0 179.021739 38.784058 130.0 150.00 172.5 205.000 320.0
range_km 276.0 354.891304 107.165068 95.0 265.00 365.0 431.250 640.0
efficiency_wh/km 276.0 200.137681 33.484801 150.0 174.00 192.0 217.000 295.0
ChargeSpeed_km/hr 276.0 4.196429 2.219032 1.0 2.00 4.0 6.000 9.0
drive 276.0 2.000000 0.882146 1.0 1.00 2.0 3.000 3.0
NumberofSeats 276.0 5.304348 1.009833 2.0 5.00 5.0 5.000 7.0
PriceinGermany 276.0 68599.975738 39872.583226 21940.0 44657.25 57195.0 71849.250 400000.0
PriceinUK 276.0 53853.342830 36782.271092 6429.0 34758.75 45477.5 61321.250 350000.0

We will create a new data frame df1 to store the values after removing the outliers.

Using (Q1 - 1.5 IQR) as the minimum cap and (Q3 + 1.5 IQR) as the max cap.

In [122]:
df.columns
Out[122]:
Index(['Brand', 'Batery_KWh', 'accln_sec', 'TopSpeed_km/h', 'range_km',
       'efficiency_wh/km', 'ChargeSpeed_km/hr', 'drive', 'NumberofSeats',
       'PriceinGermany', 'PriceinUK'],
      dtype='object')
In [123]:
# Graphically analysing the distribution of data
In [124]:
plt.figure(figsize=(16,5))
plt.subplot(1,3,1)
sns.distplot(df['Batery_KWh'])

plt.subplot(1,3,2)
sns.distplot(df['accln_sec'])

plt.subplot(1,3,3)
sns.distplot(df['TopSpeed_km/h'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3526774868.py:3: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3526774868.py:6: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3526774868.py:9: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[124]:
<Axes: xlabel='TopSpeed_km/h', ylabel='Density'>
In [125]:
plt.figure(figsize=(16,5))

plt.subplot(1,3,1)
sns.distplot(df['range_km'])

plt.subplot(1,3,2)
sns.distplot(df['efficiency_wh/km'])

plt.subplot(1,3,3)
sns.distplot(df['ChargeSpeed_km/hr'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\918168938.py:4: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\918168938.py:7: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\918168938.py:10: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[125]:
<Axes: xlabel='ChargeSpeed_km/hr', ylabel='Density'>
In [ ]:
In [126]:
plt.figure(figsize=(16,5))

plt.subplot(1,3,1)
sns.distplot(df['NumberofSeats'])

plt.subplot(1,3,2)
sns.distplot(df['PriceinGermany'])

plt.subplot(1,3,3)
sns.distplot(df['PriceinUK'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\122704065.py:4: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\122704065.py:7: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\122704065.py:10: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[126]:
<Axes: xlabel='PriceinUK', ylabel='Density'>
In [127]:
plt.figure(figsize=(16,10))

plt.subplot(3,2,1)
sns.distplot(df['drive'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3983977722.py:4: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[127]:
<Axes: xlabel='drive', ylabel='Density'>

Upon careful examination we found that the data is not normally distributed

In [128]:
df1=df #creating a new dataframe

Removing the outliers¶

In [129]:
# Batery_KWh
q1=df1['Batery_KWh'].quantile(0.25)
q3=df1['Batery_KWh'].quantile(0.75)
iqr=q3-q1
print(iqr)
ll=q1-1.5*iqr
ul=q3+1.5*iqr
df1=df1[~((df['Batery_KWh']<ll) | (df1['Batery_KWh']>ul))]
df1.shape

31.700000000000003
Out[129]:
(276, 11)
In [130]:
# accln_sec
q1=df1['accln_sec'].quantile(0.25)
q3=df1['accln_sec'].quantile(0.75)
iqr=q3-q1
print(iqr)
ll=q1-1.5*iqr
ul=q3+1.5*iqr
df1=df1[~((df['accln_sec']<ll) | (df1['accln_sec']>ul))]
df1.shape

4.325000000000001
Out[130]:
(276, 11)
In [131]:
# TopSpeed_km/h
q1=df1['TopSpeed_km/h'].quantile(0.25)
q3=df1['TopSpeed_km/h'].quantile(0.75)
iqr=q3-q1
print(iqr)
ll=q1-1.5*iqr
ul=q3+1.5*iqr
df1=df1[~((df['TopSpeed_km/h']<ll) | (df1['TopSpeed_km/h']>ul))]
df1.shape

55.0
Out[131]:
(275, 11)
In [132]:
# range_km
q1=df1['range_km'].quantile(0.25)
q3=df1['range_km'].quantile(0.75)
iqr=q3-q1
print(iqr)
ll=q1-1.5*iqr
ul=q3+1.5*iqr
df1=df1[~((df['range_km']<ll) | (df1['range_km']>ul))]
df1.shape

167.5

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\447240164.py:8: UserWarning:

Boolean Series key will be reindexed to match DataFrame index.
Out[132]:
(275, 11)
In [133]:
# efficiency_wh/km
q1=df1['efficiency_wh/km'].quantile(0.25)
q3=df1['efficiency_wh/km'].quantile(0.75)
iqr=q3-q1
print(iqr)
ll=q1-1.5*iqr
ul=q3+1.5*iqr
df1=df1[~((df['efficiency_wh/km']<ll) | (df1['efficiency_wh/km']>ul))]
df1.shape

43.0

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\2767794084.py:8: UserWarning:

Boolean Series key will be reindexed to match DataFrame index.
Out[133]:
(267, 11)
In [134]:
# ChargeSpeed_km/hr
q1=df1['ChargeSpeed_km/hr'].quantile(0.25)
q3=df1['ChargeSpeed_km/hr'].quantile(0.75)
iqr=q3-q1
print(iqr)
ll=q1-1.5*iqr
ul=q3+1.5*iqr
df1=df1[~((df['ChargeSpeed_km/hr']<ll) | (df1['ChargeSpeed_km/hr']>ul))]
df1.shape

4.0

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3507654871.py:8: UserWarning:

Boolean Series key will be reindexed to match DataFrame index.
Out[134]:
(267, 11)
In [135]:
# PriceinGermany
q1=df1['PriceinGermany'].quantile(0.25)
q3=df1['PriceinGermany'].quantile(0.75)
iqr=q3-q1
print(iqr)
ll=q1-1.5*iqr
ul=q3+1.5*iqr
df1=df1[~((df['PriceinGermany']<ll) | (df1['PriceinGermany']>ul))]
df1.shape

27995.0

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\2758903175.py:8: UserWarning:

Boolean Series key will be reindexed to match DataFrame index.
Out[135]:
(237, 11)
In [136]:
# PriceinUK
q1=df1['PriceinUK'].quantile(0.25)
q3=df1['PriceinUK'].quantile(0.75)
iqr=q3-q1
print(iqr)
ll=q1-1.5*iqr
ul=q3+1.5*iqr
df1=df1[~((df['PriceinUK']<ll) | (df1['PriceinUK']>ul))]
df1.shape

23010.0

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\701170901.py:8: UserWarning:

Boolean Series key will be reindexed to match DataFrame index.
Out[136]:
(228, 11)
In [137]:
print('Total number of rows removed', df.shape[0] - df1.shape[0])
print('There were originally', df.shape[0] ,'rows present.')

Total number of rows removed 48
There were originally 276 rows present.

Comparing the graphs of data frame df and data frame df1¶

In [138]:
plt.figure(figsize=(12,3))
plt.subplot(1,2,1)
sns.distplot(df['Batery_KWh'])

plt.subplot(1,2,2)
sns.distplot(df1['Batery_KWh'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\1827589121.py:3: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\1827589121.py:6: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[138]:
<Axes: xlabel='Batery_KWh', ylabel='Density'>
In [139]:
plt.figure(figsize=(10,6))
plt.subplot(2,2,1)
sns.distplot(df['accln_sec'])

plt.subplot(2,2,2)
sns.distplot(df1['accln_sec'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\1795226719.py:3: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\1795226719.py:6: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[139]:
<Axes: xlabel='accln_sec', ylabel='Density'>
In [140]:
plt.figure(figsize=(10,6))
plt.subplot(2,2,3)
sns.distplot(df['TopSpeed_km/h'])

plt.subplot(2,2,4)
sns.distplot(df1['TopSpeed_km/h'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3652080977.py:3: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3652080977.py:6: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[140]:
<Axes: xlabel='TopSpeed_km/h', ylabel='Density'>
In [141]:
plt.figure(figsize=(10,6))
plt.subplot(2,2,1)
sns.distplot(df['range_km'])

plt.subplot(2,2,2)
sns.distplot(df1['range_km'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\231585420.py:3: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\231585420.py:6: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[141]:
<Axes: xlabel='range_km', ylabel='Density'>
In [142]:
plt.figure(figsize=(10,6))
plt.subplot(2,2,1)
sns.distplot(df['efficiency_wh/km'])

plt.subplot(2,2,2)
sns.distplot(df1['efficiency_wh/km'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3096015979.py:3: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3096015979.py:6: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[142]:
<Axes: xlabel='efficiency_wh/km', ylabel='Density'>
In [143]:
plt.figure(figsize=(10,6))
plt.subplot(2,2,1)
sns.distplot(df['ChargeSpeed_km/hr'])

plt.subplot(2,2,2)
sns.distplot(df1['ChargeSpeed_km/hr'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\426870677.py:3: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\426870677.py:6: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[143]:
<Axes: xlabel='ChargeSpeed_km/hr', ylabel='Density'>
In [144]:
plt.figure(figsize=(10,6))
plt.subplot(2,2,1)
sns.distplot(df['drive'])

plt.subplot(2,2,2)
sns.distplot(df1['drive'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\1651637993.py:3: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\1651637993.py:6: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[144]:
<Axes: xlabel='drive', ylabel='Density'>
In [145]:
plt.figure(figsize=(10,6))
plt.subplot(2,2,1)
sns.distplot(df['PriceinGermany'])

plt.subplot(2,2,2)
sns.distplot(df1['PriceinGermany'])

C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3328977471.py:3: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751


C:\Users\prati\AppData\Local\Temp\ipykernel_2916\3328977471.py:6: UserWarning:



`distplot` is a deprecated function and will be removed in seaborn v0.14.0.

Please adapt your code to use either `displot` (a figure-level function with
similar flexibility) or `histplot` (an axes-level function for histograms).

For a guide to updating your code to use the new functions, please see
https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751
Out[145]:
<Axes: xlabel='PriceinGermany', ylabel='Density'>

On comparisn we found that the data is taking the shape of bell curve after the outliers has been removed.

In [146]:
df1.describe().T
Out[146]:
count mean std min 25% 50% 75% max
Batery_KWh 228.0 65.346930 18.823079 16.7 50.800 68.000000 77.00 108.4
accln_sec 228.0 8.183333 2.805954 3.3 6.175 7.850000 9.75 14.3
TopSpeed_km/h 228.0 169.587719 31.641803 130.0 150.000 160.000000 185.00 261.0
range_km 228.0 338.618421 101.666506 95.0 260.000 347.500000 410.00 615.0
efficiency_wh/km 228.0 195.907895 31.126505 150.0 172.750 188.500000 207.00 262.0
ChargeSpeed_km/hr 228.0 3.987782 2.021628 1.0 2.000 4.000000 5.00 9.0
drive 228.0 1.864035 0.846826 1.0 1.000 2.000000 3.00 3.0
NumberofSeats 228.0 5.298246 0.965756 2.0 5.000 5.000000 5.00 7.0
PriceinGermany 228.0 55550.167999 17346.996616 21940.0 42330.000 53707.000000 63300.00 113359.0
PriceinUK 228.0 42573.275923 19033.912698 6429.0 31870.000 44131.150585 53680.00 89500.0
In [ ]:

Importing ML libraries¶

In [147]:
pip install scikit-learn --upgrade

Requirement already satisfied: scikit-learn in c:\users\prati\anaconda3\lib\site-packages (1.3.2)Note: you may need to restart the kernel to use updated packages.

[notice] A new release of pip is available: 23.2.1 -> 23.3.1
[notice] To update, run: python.exe -m pip install --upgrade pip

Requirement already satisfied: numpy<2.0,>=1.17.3 in c:\users\prati\anaconda3\lib\site-packages (from scikit-learn) (1.23.5)
Requirement already satisfied: scipy>=1.5.0 in c:\users\prati\anaconda3\lib\site-packages (from scikit-learn) (1.10.0)
Requirement already satisfied: joblib>=1.1.1 in c:\users\prati\anaconda3\lib\site-packages (from scikit-learn) (1.1.1)
Requirement already satisfied: threadpoolctl>=2.0.0 in c:\users\prati\anaconda3\lib\site-packages (from scikit-learn) (2.2.0)
In [148]:
pip install  xgboost

Requirement already satisfied: xgboost in c:\users\prati\anaconda3\lib\site-packages (2.0.1)
Requirement already satisfied: numpy in c:\users\prati\anaconda3\lib\site-packages (from xgboost) (1.23.5)
Requirement already satisfied: scipy in c:\users\prati\anaconda3\lib\site-packages (from xgboost) (1.10.0)
Note: you may need to restart the kernel to use updated packages.

[notice] A new release of pip is available: 23.2.1 -> 23.3.1
[notice] To update, run: python.exe -m pip install --upgrade pip
In [149]:
from sklearn.model_selection import train_test_split
from sklearn.pipeline import Pipeline
from sklearn.linear_model import LogisticRegression, LinearRegression
from sklearn.ensemble import RandomForestRegressor
from sklearn.tree import DecisionTreeRegressor
from xgboost import XGBRegressor
from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score, f1_score

Creating Feature Matrix x and storing target value in y¶

NOTE- for accuracy and efficiency only the price of German market has been considered

In [150]:
df1.columns
Out[150]:
Index(['Brand', 'Batery_KWh', 'accln_sec', 'TopSpeed_km/h', 'range_km',
       'efficiency_wh/km', 'ChargeSpeed_km/hr', 'drive', 'NumberofSeats',
       'PriceinGermany', 'PriceinUK'],
      dtype='object')
In [151]:
# Define features and target variables
X = df1.drop(columns=['PriceinGermany', 'PriceinUK'])
y = df1[['PriceinGermany']]
In [152]:
X = X.drop(columns=['Brand']) #brand value may be a factor affecting price but it is not the primary factor hence can be ignored here
In [219]:
# train test split
#since data is not very large, only 20 % is taken as test data
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=92)
In [204]:
print(X_train.shape,y_train.shape)

(182, 8) (182, 1)
In [205]:
print(X_test.shape,y_test.shape)

(46, 8) (46, 1)

Estimate the coefficients for each input feature.¶

In [206]:
import statsmodels.formula.api as smf
import statsmodels.api as sm
model=sm.OLS(y_train,X_train).fit()
model.summary()
Out[206]:
OLS Regression Results
Dep. Variable: PriceinGermany R-squared (uncentered): 0.971
Model: OLS Adj. R-squared (uncentered): 0.970
Method: Least Squares F-statistic: 726.3
Date: Sat, 28 Oct 2023 Prob (F-statistic): 2.29e-129
Time: 00:12:15 Log-Likelihood: -1932.9
No. Observations: 182 AIC: 3882.
Df Residuals: 174 BIC: 3907.
Df Model: 8
Covariance Type: nonrobust
coef std err t P>|t| [0.025 0.975]
Batery_KWh 1210.5890 213.590 5.668 0.000 789.028 1632.150
accln_sec -162.0714 675.417 -0.240 0.811 -1495.136 1170.993
TopSpeed_km/h 283.0061 47.827 5.917 0.000 188.611 377.401
range_km -161.0347 41.262 -3.903 0.000 -242.474 -79.596
efficiency_wh/km -4.9222 81.163 -0.061 0.952 -165.112 155.268
ChargeSpeed_km/hr -790.3493 513.228 -1.540 0.125 -1803.303 222.605
drive -1439.0378 1563.245 -0.921 0.359 -4524.401 1646.325
NumberofSeats -1706.8887 1432.278 -1.192 0.235 -4533.764 1119.987
Omnibus: 40.606 Durbin-Watson: 2.274
Prob(Omnibus): 0.000 Jarque-Bera (JB): 102.789
Skew: 0.943 Prob(JB): 4.78e-23
Kurtosis: 6.162 Cond. No. 1.02e+03


Notes:
[1] R² is computed without centering (uncentered) since the model does not contain a constant.
[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[3] The condition number is large, 1.02e+03. This might indicate that there are
strong multicollinearity or other numerical problems.
In [ ]:
In [207]:
# Regression models
random_forest = RandomForestRegressor()
decision_tree = DecisionTreeRegressor()
xg_boost =  XGBRegressor()
lin_reg  =  LinearRegression()

model training and validation¶

In [208]:
mypipeline= [random_forest, decision_tree, xg_boost, lin_reg]

# training the pipelines
for pipe in mypipeline:
    pipe.fit(X_train,y_train)

C:\Users\prati\anaconda3\lib\site-packages\sklearn\base.py:1152: DataConversionWarning:

A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().
In [209]:
PipelineDict = {0: 'Random Forest', 1: 'Decision Tree', 2:'XG Boost', 3: 'Linear Regression'}
In [220]:
for i,model in enumerate(mypipeline):
    print('{} TestAccuracy:{}'.format
         (PipelineDict[i],model.score(X_test,y_test)))

Random Forest TestAccuracy:0.8941580637830726
Decision Tree TestAccuracy:0.9178649563850655
XG Boost TestAccuracy:0.9160108790498399
Linear Regression TestAccuracy:0.7374683649530456
In [221]:
accuracy=0
classifier=0
pipeline=""
In [222]:
for i, model in enumerate(mypipeline):
    if model.score(X_test,y_test)>accuracy:
        accuracy=model.score(X_test,y_test)
        pipeline= model
        classifier=i

print('classifier with best accuracy:{}'.format(PipelineDict[classifier]))

classifier with best accuracy:Decision Tree

Since decision_tree has the highest accuracy we will proceed with linear regression model for calculating error and F1 score

Errors for Linear Regression¶

In [223]:
decision_tree.score(X_test,y_test)
Out[223]:
0.9178649563850655
In [224]:
predictions = decision_tree.predict(X_test) #prediction
predictions
Out[224]:
array([ 59200.        ,  36400.        ,  67300.        ,  44668.        ,
        70800.        ,  47595.        ,  95990.        , 108867.        ,
        54430.        ,  53000.        ,  57940.        ,  57000.        ,
        59580.        ,  35700.        ,  42644.66666667,  94123.        ,
        69950.        ,  66069.        ,  33240.        ,  69950.        ,
        33240.        ,  61100.        ,  37900.        ,  39682.5       ,
        55900.        ,  47564.875     ,  47564.875     ,  40000.        ,
        37840.        ,  95074.        ,  59128.        ,  55980.        ,
        36790.        ,  42145.        ,  40846.66666667,  55162.4       ,
        36790.        ,  64530.        ,  54430.        ,  59950.        ,
        42900.        ,  36840.        ,  89351.        ,  40650.        ,
        54000.        ,  40150.        ])
In [225]:
# Calculate MSE
mse = mean_squared_error(y_test, predictions)
mse
Out[225]:
25085634.753916066
In [226]:
# Calculate MAE
mae = mean_absolute_error(y_test, predictions)
mae
Out[226]:
2215.8916666666664
In [167]:
# Calculate R2 score
r2 = r2_score(y_test, lin_predictions)
r2
Out[167]:
0.5527793683892224
In [227]:
import math

# Calculate RMSE
rmse = math.sqrt(mse)
rmse
Out[227]:
5008.5561546134295
In [ ]:

Predicting values¶

In [171]:
df1.head()
Out[171]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
3 Renault 52.0 11.4 135 315 165 2.0 1 5 36840.000 9570.000
4 Audi 52.0 9.0 160 285 182 3.0 2 5 70905.625 56860.625
5 Tesla 75.0 3.7 250 415 181 6.0 3 5 63667.000 59990.000
6 Porsche 83.7 5.1 220 425 197 1.0 3 4 98514.000 84500.000
7 Renault 52.0 9.5 140 310 168 2.0 1 5 37840.000 29995.000
In [230]:
y_predict = decision_tree.predict([[52.0,11.4,135,315,165,2.0,1,5]]) #taking the 1st value from df1.head()
y_predict 

C:\Users\prati\anaconda3\lib\site-packages\sklearn\base.py:465: UserWarning:

X does not have valid feature names, but DecisionTreeRegressor was fitted with feature names
Out[230]:
array([36840.])
In [173]:
df1.NumberofSeats.value_counts()
Out[173]:
5    156
7     48
4     22
2      2
Name: NumberofSeats, dtype: int64

Using case1 and cas2 for prediction which were created while cleaning data¶

In [174]:
case1.head()
Out[174]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
56 JAC 39.0 12.0 132 225 173 1.0 Front Wheel Drive 5 0.0 0.0
214 Hongqi 76.5 6.5 200 305 251 4.0 All Wheel Drive 5 0.0 0.0
215 Hongqi 90.0 4.9 200 355 254 4.0 All Wheel Drive 5 0.0 0.0
271 Seres 51.0 8.9 155 260 196 2.0 Front Wheel Drive 5 0.0 0.0
287 XPENG 82.7 6.7 200 500 165 7.0 Rear Wheel Drive 5 0.0 0.0
In [231]:
y_predict = decision_tree.predict([[39.0,12.0,132,225,173,1.0,1,5]]) #taking the 1st value from case1.head()
y_predict 

C:\Users\prati\anaconda3\lib\site-packages\sklearn\base.py:465: UserWarning:

X does not have valid feature names, but DecisionTreeRegressor was fitted with feature names
Out[231]:
array([50832.22222222])
In [176]:
case2.head()
Out[176]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
0 Lucid 118.0 2.7 270 645 183 8.0 All Wheel Drive 5 218000.0 0.0
31 Lucid 112.0 3.2 270 665 168 8.0 All Wheel Drive 5 174500.0 0.0
32 Lucid 88.0 3.4 250 550 160 9.0 All Wheel Drive 5 120000.0 0.0
33 Lucid 88.0 4.2 200 560 157 9.0 Rear Wheel Drive 5 100000.0 0.0
122 SsangYong 56.0 8.5 156 290 193 3.0 Front Wheel Drive 5 40490.0 0.0
In [232]:
y_predict = decision_tree.predict([[118.0,2.7,270,645,183,8.0,3,5]]) #taking the 1st value from case2.head()
y_predict 

C:\Users\prati\anaconda3\lib\site-packages\sklearn\base.py:465: UserWarning:

X does not have valid feature names, but DecisionTreeRegressor was fitted with feature names
Out[232]:
array([113359.])
In [ ]:

Applications of the ML model¶

This ML model can help:

  1. Car buyers make informed decisions by comparing different cars and finding the one that best meets their needs and budget.
  2. Car dealers set competitive prices based on car features and market conditions.
  3. R&D engineers make informed decisions when designing cars.
  4. Banks and lenders make more informed lending decisions to reduce the risk of defaults.
  5. Governments develop transportation policies that benefit consumers and the environment.
In [ ]:

Converting the final data frame to CSV¶

In [194]:
df1.head()
Out[194]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
3 Renault 52.0 11.4 135 315 165 2.0 1 5 36840.000 9570.000
4 Audi 52.0 9.0 160 285 182 3.0 2 5 70905.625 56860.625
5 Tesla 75.0 3.7 250 415 181 6.0 3 5 63667.000 59990.000
6 Porsche 83.7 5.1 220 425 197 1.0 3 4 98514.000 84500.000
7 Renault 52.0 9.5 140 310 168 2.0 1 5 37840.000 29995.000
In [195]:
# Define the reverse mapping dictionary
reverse_drive_mapping = {1: 'Front Wheel Drive', 2: 'Rear Wheel Drive', 3: 'All Wheel Drive'}

# Use the map function to restore the original values
df1['drive'] = df1['drive'].map(reverse_drive_mapping)
In [196]:
df1.head()
Out[196]:
Brand Batery_KWh accln_sec TopSpeed_km/h range_km efficiency_wh/km ChargeSpeed_km/hr drive NumberofSeats PriceinGermany PriceinUK
3 Renault 52.0 11.4 135 315 165 2.0 Front Wheel Drive 5 36840.000 9570.000
4 Audi 52.0 9.0 160 285 182 3.0 Rear Wheel Drive 5 70905.625 56860.625
5 Tesla 75.0 3.7 250 415 181 6.0 All Wheel Drive 5 63667.000 59990.000
6 Porsche 83.7 5.1 220 425 197 1.0 All Wheel Drive 4 98514.000 84500.000
7 Renault 52.0 9.5 140 310 168 2.0 Front Wheel Drive 5 37840.000 29995.000
In [197]:
df1.to_csv("S:\BIA Capstone/EV_dataset_final.csv")
In [ ]: